by Katy Spink, Ph.D.,
Dark Horse Consulting Group’s COO & MP
Katy Spink Presents at an Internal DHC Retreat in 2018
Full disclosure: my CGT career began just over 20 years ago when I worked with pluripotent stem cells at Geron. I’ve always been a believer in the potential of that technology to deliver highly effective therapies with significant advantages in scalability, cost effectiveness and manufacturing consistency. However, it is difficult to deny that the promise of this technology has been a long time coming. For at least two decades, it has been perceived as being ‘just over the horizon.’ One could be forgiven for wondering when exactly that promise would come to fruition.
Recent developments in the space, though, have me wondering whether we may finally be on the verge of developing highly effective cell therapy products derived from pluripotent cells.
In June of this year we saw an announcement that all six evaluable patients dosed in a phase 1/2 trial of Vertex’s VX-880 diabetes cell therapy demonstrated endogenous insulin secretion, improved HbA1c, improved time-in-range on continuous glucose monitoring, and reduction or elimination of external insulin control. Additionally, the two patients who were beyond the 12-month primary follow-up period both met the primary efficacy endpoints of elimination of severe hypoglycemic events (SHEs) between 3- and 12-months post infusion (despite a baseline of recurrent SHEs in the year before treatment), with a reduction of Hb1Ac. Both patients further demonstrated insulin independence, with over 95% time-in-range on continuous glucose monitoring (well above the ADA recommended target of ≥70%).
Although the patient numbers evaluated so far are small and the timepoints evaluated are relatively short, for those of us who have followed the field’s efforts to develop functional islet cells, this is truly a remarkable milestone that represents a sea change from the progress of the field over the prior 20+ years.
Although I’m not a clinician, I don’t think I’m going out on a limb to say that insulin independence with elimination of SHEs and well above the ADA recommended target for glucose control, in a population with impaired hypoglycemic awareness and a history of recurrent SHEs, would be considered a highly clinically significant result…if it holds up in a larger population and for longer periods of time.
This news was followed in August by an announcement of promising clinical data for a second pluripotent stem cell derived therapy: BlueRock/Bayer’s BRT-DA01 (bemdaneprocel; human embryonic stem cell-derived dopaminergic (DA) neuron precursors) in Parkinson’s disease. Although efficacy endpoints in Parkinson’s (as with many neurological diseases) are somewhat less easily interpreted than in diabetes, the exploratory clinical endpoints evaluated in the study appeared to show great promise, with dose-dependent improvements in both of the pre-specified exploratory efficacy endpoints.
Furthermore (and let me repeat the important caveat that I’m not a clinician!), my literature research suggested that the observed differences of 13 points (high dose cohort) and 7.6 points (low dose cohort) on the Unified Parkinson’s Disease Rating Scale Motor Score (UPDRS Part III) would likely be considered quite clinically significant by the physician community. (Reference both https://www.ncbi.nlm.nih.gov/pmc/articles/PMC6336372/ and https://jamanetwork.com/journals/jamaneurology/fullarticle/799064.)
So, why the sudden leap forward, more than a decade after Geron initiated the first clinical trial of a pluripotent stem cell derived cell therapy product? It’s my view that our ability to demonstrate efficacy of pluripotent cell derived therapies has historically been hindered by two primary factors: cellular potency, and cell survival and engraftment. Let’s explore where these two products—and the field as a whole—stand relative to those historical challenges.
A key challenge for the field has been to identify the precise cell type (or mix of cell types) that represents the optimal replacement therapy for any specific condition. In other words, when you have the wealth of opportunity that comes with the ability to (at least theoretically) make any cell type of the human body, which cell or mix of cells should you make? And while it has been dogma in the field for many years that progenitor cells are more likely to survive cell transplant, engraft appropriately, and divide to repopulate the injury site, precisely what stage of cellular maturity is ideal to balance survival and engraftment with maturation and function?
For example, when I was Program Director for Geron’s GRNCM1 (hESC-derived cardiomyocytes) program for heart failure, the conventional wisdom was generally that a process that derived a higher percentage of cardiomyocytes was preferred. But, given that cardiomyocytes account for only about 30% of the total cells in the human heart, we often debated whether having some naturally occurring supportive cell types such as fibroblasts and smooth muscle cells in the product might be desirable for product engraftment and function. In fact, tissue engineering studies in recent years have suggested this might be the case, demonstrating improved contractile and electrophysiological properties in constructs with mixed cellular compositions (although the implications of this for the ideal composition of an injected cellular suspension is still unclear).
Turning to the Vertex and BlueRock products, it seems highly likely that advances in cellular potency (i.e., delivering a more functional cell) played a role in the recent successes of both products. While previous trials seeking to restore pancreatic islet function delivered pancreatic precursor cells requiring significant in vivo maturation, VX-880 is delivered as a fully differentiated insulin-producing cell, based on the groundbreaking research of the Melton lab at Harvard to develop protocols to differentiate fully functional islets in vitro. Similarly, BlueRock collaborated with leading researchers Lorenz Studer and Vivian Tabar of MSKCC for years before entering the clinic, during which time they were presumably improving on the published protocols for DA neuron production from those investigators.
So, what does this mean for the future of the field of PSC-derived therapies? My view is it’s a sign of good progress, but not a dramatic leap forward for the whole field. While there may be some common learnings across cell types, more likely we’ll need to figure out the optimal phenotype on a case-by-case basis for each indication, meaning progress is more likely to be made one indication at a time than by leaps and bounds for the whole field at once.
A host of PSC-derived T-cell therapies in the pipeline provide another example of how advances in our understanding of optimal cellular potency are informing the development of PSC-based therapies. Autologous CAR-T programs have taught us a lot about how to develop potent immunotherapy products, including how to supercharge cytotoxicity with optimal CAR designs, transgene armoring, and which T-cell subtypes to target in the product mix for improved persistence. Recent advances in T-cell differentiation methods have made production of functional T-cells from PSCs feasible. The final hurdle will be figuring out how to optimize persistence of the allogeneic product in the host, which brings me to our second historical challenge.
Cell survival and engraftment:
In order for most cell therapies to be effective, the cells need to survive for long enough to have the desired effect. We’ve seen this with autologous CAR-T therapies, where studies have demonstrated a link between cellular persistence and durable remissions.
For PSC-derived therapies, the bar is often even higher because frequently the outcomes being sought may require not just cell survival but actual functional engraftment of the delivered cells into the target tissue, including physiologically appropriate integration and coupling with host cells. Since PSC-derived therapies are usually allogeneic and may often be transplanted into challenging environments for cell survival due to inflammation, immunoreactivity, low oxygen, reduced vascularization, and other factors (depending on the condition a severity of disease state being treated), achieving long term graft persistence with appropriate functional integration has historically been a key barrier to success for the field.
For VX-880, this challenge is arguably a bit more tractable, given that pancreatic islet cells do not require integration in the pancreas to function. However, previous attempts by other sponsors to protect cells from immune rejection via encapsulation in a device have proven more challenging than many anticipated. For its first-in-human study, Vertex took the far more straightforward approach of following the proof of concept established by the Edmonton Protocol, with delivery via hepatic portal vein infusion followed by chronic immunosuppression. In my opinion, this was a smart choice, as it provided an opportunity to test cellular function of their optimized differentiation protocol using a method that had already been shown to be supportive of cell survival and function for delivery of cadaveric islets. Follow-on approaches being pursued by Vertex and others include both encapsulation in an immunoprotective device, and use of gene editing to create hypoimmune islets for transplantation.
Similarly, for BRT-DA01, it appears that cell survival was conferred solely through immunosuppression, as opposed to the use of any newer technologies such as using gene edited hypoimmune cells to improve survival and engraftment. Of note, it does appear that BlueRock used a more aggressive two-drug immunosuppressive regimen of prednisone plus tacrolimus, rather than the prior precedent of tacrolimus alone in Geron’s spinal cord injury trial. Among the next generation approaches to improving cell survival and functional engraftment of PSC-derived therapies in the pipeline for Parkinson’s Disease are autologous approaches, and the use of microtissue grafts to improve graft integration.
Looking at the pipeline more broadly, a range of gene-editing approaches are in development to establish hypoimmune PSC-based therapies that can survive, engraft and function in an allogeneic setting, ideally without the need for long-term immunosuppression. While the optimal set of edits to enable hypoimmunity are not yet fully clear, the many parallel efforts that are ongoing against this goal give me hope that an answer is near. What is clear is that PSCs represent an ideal platform for the application of this technology, given the ability to edit cells once upstream of a Master Cell Bank (MCB) and then supply an entire product lifecycle from that MCB, conferring significant advantages in scalability, reproducibility and safety for a multiple-gene edited product.
So, almost 15 years after PSC-derived therapies first entered the clinic, are we finally on the cusp of realizing their potential? I would argue that these two recent early trials suggest we are. Perhaps most encouragingly, it appears that these recent successes were enabled primarily by the relatively straightforward approach of optimizing cellular potency and immunosuppression, leaving significant room for further improvement from cutting edge approaches still emerging in the pipeline.
The field of cell therapy has always been an exciting ride, but it may be about to get even more so. Are we there yet? Buckle up, and let’s see.